Clinical characteristics and factors associated with severe acute respiratory infection and influenza among children in Jingzhou, China

Background Influenza is an important cause of respiratory illness in children, but data are limited on hospitalized children with laboratory‐confirmed influenza in China. Methods We conducted active surveillance for severe acute respiratory infection (SARI; fever and at least one sign or symptom of acute respiratory illness) among hospitalized pediatric patients in Jingzhou, Hubei Province, from April 2010 to April 2012. Data were collected from enrolled SARI patients on demographics, underlying health conditions, clinical course of illness, and outcomes. Nasal swabs were collected and tested for influenza viruses by reverse transcription polymerase chain reaction. We described the clinical and epidemiological characteristics of children with influenza and analyzed the association between potential risk factors and SARI patients with influenza. Results During the study period, 15 354 children aged <15 years with signs and symptoms of SARI were enrolled at hospital admission. severe acute respiratory infection patients aged 5–15 years with confirmed influenza (H3N2) infection were more likely than children without influenza to have radiographic diagnosis of pneumonia (11/31, 36% vs 15/105, 14%. P<.05). Only 16% (1116/7145) of enrolled patients had received seasonal trivalent influenza vaccination within 12 months of hospital admission. Non‐vaccinated influenza cases were more likely than vaccinated influenza cases to have pneumonia (31/133, 23% vs 37/256, 15%, P<.05). severe acute respiratory infection cases aged 5–15 years diagnosed with influenza were also more likely to have a household member who smoked cigarettes compared with SARI cases without a smoking household member (54/208, 26% vs 158/960, 16%, P<.05). Conclusions Influenza A (H3N2) virus infection was an important contributor to pneumonia requiring hospitalization. Our results highlight the importance of surveillance in identifying factors for influenza hospitalization, monitoring adherence to influenza prevention and treatment strategies, and evaluating the disease burden among hospitalized pediatric SARI patients. Influenza vaccination promotion should target children.


| BACKGROUND
Influenza is an important vaccine-preventable infectious disease that causes significant morbidity worldwide in persons of all ages.
Studies conducted in North America, Europe, and Asia have documented that children have particularly high rates of morbidity and influenza-related complications, particularly those aged <2 years. [1][2][3] Although effectiveness varies by season through the interaction of viral and host factors, annual influenza vaccination remains the most effective method for preventing seasonal influenza and related complications. [4][5][6][7] In the United States, annual influenza vaccination is recommended for all persons aged ≥6 months of age. 8 10 We conducted active SARI surveillance in central China to analyze the clinical and epidemiological characteristic of influenza-associated SARI cases, including pneumonia, among pediatric patients aged <15 years.

| METHODS
Surveillance was conducted in Jingzhou, Hubei Province, in three general hospitals and one pediatric hospital as previously described. 10 All patients admitted to a surveillance hospital were screened by nurses and physicians for SARI and were considered eligible if the SARI case definition was met within 24 hours of hospital admission. Briefly, we collected demographic, clinical, and outcome data and performed influenza testing for all pediatric patients (aged <15 years) hospitalized for SARI to characterize the epidemiology of severe influenza in children.

| Case definitions
Children <15 years of age met the case definition for SARI if they had measured elevated temperature (rectal or axillary) ≥37.3°C and at least one sign or symptom of acute respiratory illness, including cough, sore throat, tachypnea, difficulty breathing, abnormal breath sounds on auscultation, sputum production, hemoptysis, chest pain, or chest radiograph consistent with pneumonia. A laboratory-confirmed influenza case was defined as a SARI patient with a documented positive result of real-time reverse transcription polymerase chain reaction (RT-PCR) assay for influenza, admitted to a participating hospital between April 5, 2010, and April 8, 2012. A pneumonia case was defined as a SARI patient with infiltrates noted on chest radiograph performed at any time during the hospitalization.

| Data collection
Physicians screened potential case patients for SARI at hospital admission and obtained verbal consent from a parent or guardian for enrollment in the study. Physicians then abstracted data from the medical records of patients meeting the SARI case definition including basic demographic information, past medical history, signs, symptoms, and radiographic results on a structured case report form. At hospital discharge, physicians were required to update the form to include data about outcome and influenza testing results.

| Laboratory
Nurses collected nasal swabs from pediatric SARI case patients within 24 hours of admission following standardized procedures and were transferred to the Jingzhou Center for Disease Control and Prevention (Jingzhou CDC). At Jingzhou CDC, laboratory confirmation of influenza virus infection was determined by real-time RT-PCR. Handling and testing of specimens has been described in detail elsewhere. 10
Descriptive statistics including frequency analysis for categorical variables, medians, and interquartile ranges (IQRs) for continuous variables were calculated. To further examine the association between potential risk factors and SARI, we first performed univariate analysis.
For multivariable logistic regression, we included variables with P<.10 in univariate analysis or those believed to be potential risk factors associated with the outcome of interest, such as "at least one tobacco smoker in household" and we conducted stepwise backwards variable selection, retaining variables with P<.05 after controlling for other risk factors in the logistic regression model. Data were analyzed with spss (v17.0; SPSS, Chicago, IL, USA).

| SARI patients
The median age of all enrolled SARI patients was 1.9 years (interquartile range [IQR], 0.9-3.3 years), 80% were aged <5 years, and 57% were male. Only 2% of SARI patients had an underlying medical condition, the most common of which was prematurity (2%).
Only 1% of influenza patient received monovalent pandemic H1N1 vaccination and 16% patient received seasonal trivalent influenza vaccination (Table 1). severe acute respiratory infection patients were admitted to the hospital a median of 2 days (IQR, 1-3 days) after illness onset. Cough (61%) and sore throat (27%) were the most frequent symptoms of SARI patients at hospital admission. A quarter of all SARI patients had high fever (temperature ≥39.0°C) at admission. The median length of hospitalization for SARI patients was 5 days (IQR, [3][4][5][6]. Of these SARI patients, 3 (0.02%) died during hospitalization, and 5 (0.03%) died within 30 days after discharge (Table 2). Of  healthy, and only 2% had at least one underlying medical condition, the most common of which were low birth weight (1%) and prematurity (1%) among children aged <2 years.

| SARI patients with and without laboratoryconfirmed influenza
Of 14 479 children who met the SARI case definition and had respiratory specimens collected during the study period, 12 705 (88%) tested negative for influenza viruses. The demographic characteristics of patients with and without influenza were similar except those without influenza had a lower median age and a younger age distribution (P<.05; Table 1). Influenza cases were more likely to have at least one tobacco smoker in the household and a sick contact with fever or respiratory symptoms than children without influenza (P<.05).
Clinical characteristics of pediatric influenza cases and children without influenza were similar. More pediatric influenza cases had cough and rhinorrhea than children without influenza (P<.05), while more SARI cases without influenza had gastrointestinal symptoms (nausea, vomiting, and diarrhea) than influenza cases (
Among influenza cases, the proportion with cough, sputum production, or Dyspnoea was higher in children with radiologic evidence of pneumonia compared with those without (P<.05; Table 3). However, children without radiologic evidence of pneumonia had higher temperatures (T>38°C) on admission than children with radiologic evidence of pneumonia, but more pneumonia cases had abnormal breath sounds on auscultation than children without pneumonia (P<.05).
Among the influenza A cases, 46 (32%) were due to A(H1N1)pdm09 and 99 (68%) were due to A(H3N2). Among pneumonia cases aged 5-15 years, the proportion associated with A(H3N2) was significantly higher than either A(H1N1)pdm09 or influenza B (Table S3-1). The P-values are comparisons between "pediatric SARI patients with confirmed influenza infection" and "pediatric SARI patients without influenza infection", and statistically significant as P<.05 c There were 875 (6%) patients without NP/OP collection. d Body mass index (BMI) was calculated for patients with available height and weight data to assess obesity using Chinese criteria (BMI greater than the cutoff values for children aged 2-17 y). BMI was not calculated in children aged <2 y. 11 e Defined as gestational age <37 wk at birth for children aged <2 y. f Defined as birth weight of a live born infant of <2500 g for children aged <2 y. g Answered "Yes" to this question, "During the past 12 mo, have you had a flu shot?" which were asked during a face-to-face interview.  Data presented as no. (%) of patients unless otherwise indicated. Denominators for testing of fewer cases than full group are indicated. Percentages may not total 100 because of rounding. Data only include patients with specimens for influenza virus testing. b The P-values are comparisons between "pediatric influenza patients with pneumonia" and "pediatric influenza patients without pneumonia", and statistically significant as P<.05. c Defined as a reduction in the circulating WBC count to <4000/μL. d Defined as a reduction in the circulating WBC count to >10 000/μL. e Defined as a reduction in the hemoglobin in grams per liter (Hb<110 g/L). f Raised erythrocyte sedimentation rate defined as exceed up limit as 20 mm/h.

| Treatment
Forty-seven percent of SARI patients had either been prescribed an antibiotic or self-administered an antibiotic before hospitalization. A higher proportion of influenza-associated SARI cases had been prescribed an antibiotic before admission than those without influenza (

| Risk factors associated with influenza
The results of univariate and multivariable analyses of potential factors associated with influenza among SARI patients are presented in Tables S1-1-S8. The proportion of SARI cases with a household member who smokes cigarettes in the home or contact with anyone with fever or respiratory symptoms was significantly associated with influenza among young children aged 6-23 months. These two factors were also significantly associated with higher odds of influenza for children aged 2-4 years and children aged 5-15 years.

| DISCUSSION
To our knowledge, this is one of the first prospective studies describ-  influenza and high inappropriate use of antibiotics and corticosteroids were identified from the study.
In 2012, the WHO Strategic Advisory Group of Experts on immunization suggested that children up to 5 years of age should be considered as a target group for annual influenza vaccination. 14 Several national guidelines recommend annual influenza immunization for children. 15 While the infections (12%) in this study were caused by influenza viruses, clinicians prescribed antibiotics for almost all SARI patients.
While we did not have data to assess the need for antibiotics in our study population, inappropriate antibiotic use is potentially harmful to the community, fostering the growth of antimicrobial-resistant organisms. 29 Reducing inappropriate antibiotic use among children in China may be challenging. Physicians may not be willing to wait for viral testing results, influenza testing is not widely available, and, even when widely available as in this study, physicians often did not receive results until at least 48 hours after specimen collection. Parents may also be less willing to accept waiting for test results before antibiotic administration. Indeed, over half of SARI patients were already treated with antibiotics before hospital admission. 30 Severe acute respiratory infection patients exposed to tobacco smoke were more likely to have flu as an etiology of SARI. Environmental tobacco smoke or secondhand smoke causes ill health and mortality in children, especially among those under 5 years of age. [31][32][33] Exposure to secondhand smoke kills approximately 100 000 people every year in China. 34 Reducing secondhand smoke exposure is one of four priorities identified by the World Health Organization for global tobacco prevention and control. China's 12th 5-Year Plan calls for smoke-free public places as part of the major national goal to increase life expectancy.
Despite the significant progress made in tobacco control in China, many children are still exposed to secondhand smoke in their home.

| CONCLUSIONS
Our results highlight the importance of surveillance in identifying fac-